About Me

My photo
Chemical engineer working in the field of bulk chemicals for e.g. plastics and energy, specifically energy efficiency and renewables.

Saturday 11 June 2011

What is an engineering skeptic anyway?

On the back of recent and dramatic failures, I have been forced not only to consider why these failures have occurred but also the situation around my own position. For the last 3 years I have been working in energy research and before that, I did a Masters in the same area. I have been involved in the good, the bad and the incompetent. But one thing has been a constant. I am sometimes described as negative. I do not agree with this point. Indeed I think that this attitude is why those failures are happening. 

My job is to take the inventions of others and apply them in the real, dirty world. That is I work with the processes and perform the tests that are necessary. I therefore do not have to imagine what the performance could look like and am very aware of the pitfalls of the technologies involved. Indeed a large part of my job is trouble shooting equipment as well as the inventions themselves.
On that point I have no emotional attachment to the invention that I may test. It can be nice, it can be pretty but I do not fall in love. In many ways I compare it to my fascination with the expression of English literature in poetry. For example Evan Boland wrote
The rest is tortoiseshell and has the reticent clear patience
of its element.
I think that this describes the heartless nature of war in juxtaposition to its gross physicality. I can certainly appreciate those lines and study them but it does not mean that I love English literature.

This is where my skeptic label comes about. The inventors tend get attached to their inventions and the processes inside which they work and reside. The shortcomings can be glossed over and often the abilities of the invention are pushed too far. An example is of cellulosic  ethanol whereby companies have being promising $1-2 gas for years. However the reality is that the EPA has had to keep pushing back the date for when it expects this product to arrive on the market. Most of this comes about from having an un-realistic view of the technology and the difficulties in producing cellulosic ethanol

The best way to introduce a technology is to treat it with a short view. Essentially you want the produce to succeed but you do not believe it will until proven otherwise. It is the guilty until proven innocent principle. If you have an emotional attachment, it will be very difficult to take this approach. I consider this to be a realistic view and not a skeptic view as many of my colleagues consider. So what is the basis of this approach.

  1. ·         Everything is only a potential breakthrough until it actually has broken into the market. The sucess of an invention is based on how much it breaks into the market.
  2. ·         Pilot, pilot, pilot. If the concept is not proven on a pilot scale, then I do not consider it a potential breakthrough. Pilot scale is operating in and around 1-10% of full scale production. For example for an invention that has a lifetime of 10 years, that means continuous testing at the process conditions for 1 year is a pilot test.
  3. ·         Repeatability and reproducibility. The invention must be reproduced multiple time by different people and different equipment. Each incarnation must be shown that have similar performance (within 10%) in the same process.
  4. ·         Predictions of how the invention will perform should be taken as ‘indications only’ until pilot scale has been proven.
So what kind of success rates can you expect for the chemical industry (e.g. a new catalyst) from lab scale to full scale implementation? Approximately 1% and certainly no higher than 10%. Typically the 10% figure relates more to the concept or overall guiding idea and the 1% represents the change or each incarnation of said idea. Why such pessimism? At lab scale the feedstock is typically clean. Heavy tars and sulphur compounds for example are often missing from the feed stream. As scale up proceeds, equipment and feedstock become more complex allowing for a more complete picture to develop. The outlook changes. It is often that few ppm of poison that can ruin your catalyst and you find out the real effects at pilot scale of said poison.

Modeling the process on the computer or by hand often makes use of assumptions and ideality. Corrosion, a common problem in refinery operations it not easy (well almost impossible) to model accurately. Even after pilot, failure can occur because long term effects are hard to predict and improvements in existing and competing technologies can render the invention less favorable. 

In short, take the short view of a new claim and invention. Dig around for more data and demand that people prove their claims. You are not skeptical; just realistic. In that way failures and unrealistic assertions can be avoided.

No comments:

Post a Comment